RAID (Redundant Array of Independent Disks) is technology which enables making redundant logical disks using arrays of physical disks. Making such logical disk unit enables redundancy and better performance of disk storage. There are a few types of RAID is sense of way redundancy is achieved and technology used.
Basically, with RAID data is spread across two or more hard disks. There is more RAID types depend on way data is organized on redundant disks. Data can be spanned, striped or mirrored through physical disks. Way the data is distributed across disks defines common RAID levels:
RAID 0 – block of data are striped through physical disks
RAID 1- data are mirrored, same data exists on physical disks used in RAID
RAID 2 – bits of data are stripped across disks with dedicated disk for parity
RAID 3 – bytes of data are stripped across disks with dedicated disk for parity
RAID 4 – blocks of data are stripped across disks with dedicated disk for parity
RAID 5 – blocks of data are striped and parity is distributed across data disks
RAID 6 – blocks of data are stripped with double distributed parity across data disks
There are combination of above RAID levels like RAID 51 (RAID 5+RAID 1) or RAID 50 or similar.
As there is levels of RAID which provides redundancy and efficiency in read/write data, there are three types of RAID depends of technology for RAID controlling: hardware, firmware and software RAID. Hardware RAID assumes existing of dedicated hardware RAID controller in machine where RAID exists. It can be hardware chip on board controlling disk units or dedicated card (e.g. PCI card in PC) for hardware control of RAID. In firmware RAID there is chip for RAID but not dedicated only for Hardware RAID controlling. It can be usual hard disk controller with firmware for RAID included. Also, firmware RAID assumes driver in OS as interface for RAID firmware. Software RAID is operating system controlled RAID. It uses CPU power for RAID function.offers this type of RAID. In this article I’ll show how to configure software RAID in Windows Server 2008.
With software RAID in Windows redundancy is made on volume level. Here can be created arrays of redundant volumes where volumes are logical disks ‘seen“ by windows. Also, system volume cannot be mirrored or striped by RAID. Reason for that is windows cannot boot from volume that is in redundant configuration. You can make redundant volumes used for data to keep integrity and redundancy for your data saved on disk.
Main condition for software RAID is to have one or more redundant physical disks on server on which you configure RAID. For example, let say we have two redundant disks. After right click on Computer, then Manage and then choose Storage->Disk management appears window on which can be seen disk resources of server:
There is one system volume and two unallocated spaces on redundant physical disks. By right click on unallocated space on one of redundant disks appears menu in which can be chosen type of redundancy we want to configure:
There are four possibilities for new volume which can be created in unallocated space: new simple volume, new spanned volume, new striped volume, new mirrored volume and new RAID-5 volume. I’ll describe all options and its characteristics.
New simple volume enables creating new logical volume on unallocated space. This option does not provide redundancy, only creating new volume.
New spanned volume option enables creating new spanned volume which means volume that is spread across more physical disks but there is no redundancy for volume, only more disk space. If one of disk fails, data on volume is lost. Spanned volumes are not RAID like type of disks.
New striped volume is volume striped across two or more physical disks. This is standard RAID 0 configuration. Data are striped but there is no redundancy only expansion of disk space.
New mirrored volume enables redundancy of data mirrored on disks. When mirrored volume is created there is copy of mirrored volume on two or more physical disks. If one of disks fails there is a full copy of data on other disk and redundancy if achieved. This is standard RAID 1.
New RAID-5 volume option is for creating RAID 5 level which is block level striping with distributed parity. For this level at least three disks are needed and because of that this option is in our example greyed..
In this example we will create stripped volume across two disks. At the beginning we have disk configuration in windows server 2008 shown on next picture:
Let say we want to create stripped volume on unallocated space on disks 1 and 2. Right click on unallocated space on disk 1 and select New striped volume. Next window appears:
In this window we will choose disk for stripped volume spreading and size of volume which depends on size of unallocated space. Let say we want to choose maximum available size for stripped volume and use disks 1 and to for volume. In that case Disk 1 should be added on left field (“selected“) by click on Disk 1 and then button Add. After that click Next. New window appears:
In this window we choose drive letter for stripped volume. Let say F. Click Next.
In this window are options for volume formatting. In this example we want to create NTFS formatted volume and that option is chosen. Click Next.
Before volume is formatted physical disks must be converted to dynamic disks what is condition for making redundancy on it. Because of that conversion should be confirmed by clicking Yes. After that volume is created and formatted which can be seen on next picture:
New stripped volume is labeled with F: (New volume). Green color on disk label indicate that volume is striped (colors on the bottom of window show volume type)
In this example we created stripped volume which is analogue to RAID 0 configuration. Same way can be used for creation other types of redundant volumes. Hope this easy demonstration will give you a confident to create your own way to protect your valuable information in Windows 2008.